Nicolas LEULIER: when AI revolutionizes industrial innovation

The evening began with a question from Jérôme GORIN, teacher-researcher in cybersecurity: "Who among you uses ChatGPT on a daily basis?" The majority of hands were raised in the Jules Verne amphitheater.

This underlined the interest of the "AI and Innovation: challenges and opportunities in R&D" conference , moderated by Nicolas LEULIER (Energy Engineering & Digital Systems, 2000, Amiens), Global Technical Director at Bostik (Arkema Group). He is responsible for new product development, industrialization, customer support and implementation.

Nicolas presented the challenges and opportunities offered by AI in the industrial sector, drawing on his experience at Bostik.

According to him, AI accelerates processes, reduces costs, limits risks and minimizes environmental impact. It facilitates experimentation, data sharing and analysis, creating an innovation ecosystem. For example, an experiment carried out in Japan can be immediately exploited in France.

However, Nicolas warns that effective AI relies on structured data. Companies produce huge volumes of scattered data (Excel, Word, PDF, etc.), which must be organized, cleaned and analyzed to produce usable results.

He points out that engineers have to adapt to changing professional expectations. Traditionally, it took five years for a chemist to become fully operational through practical experience. Today, young engineers are more mobile and want to evolve quickly, often changing jobs after 2-3 years. AI meets this challenge by capturing and redistributing acquired knowledge, thus consolidating collective knowledge. It also enables engineers to acquire skills more quickly, provided they use it wisely.

Nicolas reminds us that AI is not limited to physical tasks: it also replaces some grey matter. However, its effectiveness depends on the quality of the data and its strategic management.

He warns against using public AI, such as ChatGPT, for sensitive information, recommending instead in-house solutions to preserve confidentiality and protect know-how.